The Ohio State University · Department of Astronomy
Computational & Agentic Scientific Practices, Epistemology, and Reasoning
AI is reshaping how science gets done. At Ohio State, CASPER works on three fronts at once—agentic systems for astronomical surveys, the computational methods that power modern inference, and the epistemics of discovery when machines do part of the thinking.
AI is changing how research gets done—not at the margins, but in the daily practice of discovery. CASPER, a new initiative at The Ohio State University, takes that shift seriously as a subject of study, working on three interconnected fronts: deploying LLMs as agents for large-scale astronomical surveys, advancing the computational and statistical methods modern science demands, and investigating the epistemic implications of AI-assisted discovery. Grounded in OSU's deep involvement in Roman, DESI, SDSS-V, and ASAS-SN, the work stays proven on real science—while returning to the harder question underneath: what does it mean to understand a phenomenon when AI assists the discovery?
Three fronts, one program
Bridging astronomy, physics, computer science, and philosophy—the practical challenges and the foundational questions, together.
OSU's astronomy program leads across cosmology, time-domain astronomy, stellar astrophysics, and instrumentation. Each strength connects naturally to an AI research direction.
Astronomy strength
Deep involvement in two cosmology flagships—DESI and the Roman Space Telescope—spanning instrument science, data validation, and analysis.
AI direction
Generative models for high-dimensional Bayesian inference and uncertainty quantification; agentic instrumentation control at scale.
CASPER is housed in the Department of Astronomy and draws on the interdisciplinary environment of the Center for Cosmology and AstroParticle Physics (CCAPP)—bridging physics, astronomy, and computation.
From practical deployment to foundational questions about scientific knowledge—each direction informs the others.
With deep involvement in Roman, DESI, SDSS-V, and ASAS-SN, CASPER builds and deploys agentic systems for survey operations—from instrumentation control to transient classification—grounding AI research in real applications.
Modern surveys generate data at unprecedented scale. We develop generative models for uncertainty quantification, multimodal foundation models for feature extraction, knowledge graphs for targeting, and RL for instrument control.
As AI transforms practice it raises real questions: what does it mean to understand a phenomenon when AI assists discovery? How do we evaluate scientific contributions in an era of automation? We explore these where philosophy of science meets practical AI.
Faculty, fellows, and students across astronomy, physics, computer science, and philosophy.
Center for Cosmology & AstroParticle Physics
The Ohio State University
Department of Philosophy
The Ohio State University
Emerging Technology Studio
The Ohio State University
Argonne National Laboratory
Center for Humanities & Technology
University of Cincinnati
Center for Cosmology & AstroParticle Physics
The Ohio State University
Department of Philosophy
The Ohio State University
Emerging Technology Studio
The Ohio State University
Argonne National Laboratory
Center for Humanities & Technology
University of Cincinnati
Center for Cosmology & AstroParticle Physics
The Ohio State University
Department of Philosophy
The Ohio State University
Emerging Technology Studio
The Ohio State University
Argonne National Laboratory
Center for Humanities & Technology
University of Cincinnati
Center for Cosmology & AstroParticle Physics
The Ohio State University
Department of Philosophy
The Ohio State University
Emerging Technology Studio
The Ohio State University
Argonne National Laboratory
Center for Humanities & Technology
University of Cincinnati
We seek postdoctoral researchers working at the intersection of AI and astronomical science. Fellows help build agentic systems for large-scale surveys and advance the computational methods modern astronomy needs.
Researchers who want to build AI systems for astronomical surveys
People who can bridge computational methods with scientific applications
Independent thinkers asking substantive questions about AI in science
Collaborators who contribute to practical, grounded research
